development/plasticity/repair stress ... · development/plasticity/repair...

Development/Plasticity/Repair

Stress-Activated Protein Kinase MKK7 Regulates AxonElongation in the Developing Cerebral Cortex

Tokiwa Yamasaki,1,3 Hiroshi Kawasaki,4,5 Satoko Arakawa,2 Kimiko Shimizu,6 Shigeomi Shimizu,2 Orly Reiner,7

Hideyuki Okano,8 Sachiko Nishina,9 Noriyuki Azuma,9 Josef M. Penninger,10 Toshiaki Katada,3 and Hiroshi Nishina1

Departments of 1Developmental and Regenerative Biology and 2Pathological Cell Biology, Medical Research Institute, Tokyo Medical and Dental University,Bunkyo-ku, Tokyo 113-8510, Japan, 3Department of Physiological Chemistry, Graduate School of Pharmaceutical Science, 4Department of Molecular andSystems Neurobiology, Graduate School of Medicine, 5Global COE Program “Comprehensive Center of Education and Research for Chemical Biology of theDiseases,” and 6Department of Biophysics and Biochemistry, Graduate School of Science, The University of Tokyo, Bunkyo-ku, Tokyo 113-0033, Japan,7Department of Molecular Genetics, Weizmann Institute of Science, Rehovot 76100, Israel, 8Department of Physiology, School of Medicine, Keio University,Shinjuku-ku, Tokyo 160-8582, Japan, 9Department of Ophthalmology, National Center for Child Health and Development, Setagaya-ku, Tokyo 157-8535,Japan, and 10IMBA, Institute of Molecular Biotechnology of the Austrian Academy of Science, 1030 Vienna, Austria

The c-Jun NH2-terminal protein kinase (JNK), which belongs to the mitogen-activated protein kinase family, plays important roles in abroad range of physiological processes. JNK is controlled by two upstream regulators, mitogen-activated protein kinase kinase (MKK) 7and MKK4. To elucidate the physiological functions of MKK7, we used Nestin-Cre to generate a novel mouse model in which the mkk7 genewas specifically deleted in the nervous system (Mkk7 flox/flox Nestin-Cre mice). These mice were indistinguishable from their controllittermates in gross appearance during embryogenesis but died immediately after birth without breathing. Histological examinationshowed that the mutants had severe defects in brain development, including enlarged ventricles, reduced striatum, and minimal axontracts. Electron microscopy revealed abnormal accumulations of filamentous structures and autophagic vacuoles in Mkk7 flox/flox Nestin-Cre brain. Further analysis showed that MKK7 deletion decreased numbers of TAG-1-expressing axons and delayed neuronal migrationin the cerebrum. Neuronal differentiation was not altered. In utero electroporation studies showed that contralateral projection of axonsby layer 2/3 neurons was impaired in the absence of MKK7. Moreover, MKK7 regulated axon elongation in a cell-autonomous manner invivo, a finding confirmed in vitro. Finally, phosphorylation levels of JNK substrates, including c-Jun, neurofilament heavy chain,microtubule-associated protein 1B, and doublecortin, were reduced in Mkk7 flox/flox Nestin-Cre brain. Our findings demonstrate that thephenotype of Mkk7 flox/flox Nestin-Cre mice differs substantially from that of Mkk4 flox/flox Nestin-Cre mice, and establish that MKK7-mediated regulation of JNK is uniquely critical for both axon elongation and radial migration in the developing brain.

IntroductionThe activation of c-Jun NH2-terminal protein kinase (JNK) re-sults in the phosphorylation of numerous important substrates,including the AP-1 transcription factor c-Jun (Hibi et al., 1993)and various microtubule-associated proteins (MAPs) (Kawauchiet al., 2003; Gdalyahu et al., 2004), that control cellular processessuch as cell growth, apoptosis, differentiation, migration, and

transformation. JNK activation in response to environmentalstresses, growth factors, hormones, and proinflammatory cyto-kines is triggered by mitogen-activated protein kinase (MAPK)kinase 4 (MKK4) and MKK7 (Davis, 2000; Chang and Karin,2001; Asaoka and Nishina, 2010). While MKK4 activates bothJNK and another MAPK, p38 (Enslen et al., 1998), MKK7 regu-lates only the JNK signaling cascade.

MKK7 modulates JNK signaling by interacting with scaffold pro-teins such as the JNK-interacting protein (JIP) 1, 2, or 3 or filamin A(Whitmarsh et al., 1998; Ito et al., 1999; Yasuda et al., 1999; Naka-gawa et al., 2010). Previously, we reported that MKK7 has an essen-tial role in liver development, in that Mkk7 total knock-out mice dieat E12.5–E13.5 with severely disorganized livers and reduced hepa-toblast numbers (Wada et al., 2004). This study also revealed thathepatoblast proliferation depends on the MKK7-JNK-c-Jun path-way. However, the functions of MKK7 in other tissues remain un-clear due to the early embryonic lethality of Mkk7 total knock-outmice.

Several lines of evidence have established the importance of JNKsignaling in the mammalian brain. First, Jnk1�/� mice display anabnormality in the maintenance of telencephalic commissures

Received March 3, 2011; revised Sept. 18, 2011; accepted Oct. 4, 2011.Author contributions: T.Y., H.K., T.K., and H.N. designed research; T.Y., H.K., S.A., and S.S. performed research;

T.Y., K.S., O.R., H.O., S.N., N.A., and J.M.P. contributed unpublished reagents/analytic tools; T.Y., H.K., S.A., S.S., H.O.,and H.N. analyzed data; T.Y., H.K., and H.N. wrote the paper.

This work was supported in part by research grants from the Ministry of Education, Culture, Sports, Science, andTechnology of Japan, the Ministry of Health, Labour, and Welfare of Japan, and the Japan Society for the Promotionof Science. We are grateful to numerous members of the Nishina, Kawasaki, and Katada laboratories for criticalreading and helpful discussions. We also thank Dr. Syu-ichi Hirai (Yokohama City University, Kanagawa, Japan) forhis generous and expert technical advice.

Correspondence should be addressed to Dr. Hiroshi Kawasaki, Department of Molecular and Systems Neurobiol-ogy, Graduate School of Medicine, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-0033, Japan, E-mailaddress: [email protected]; or Dr. Hiroshi Nishina, Department of Developmental and Regenerative Biol-ogy, Medical Research Institute, Tokyo Medical and Dental University, 1-5-45 Yushima, Bunkyo-ku, Tokyo 113-8510,Japan, E-mail address: [email protected].

DOI:10.1523/JNEUROSCI.1111-11.2011Copyright © 2011 the authors 0270-6474/11/3116872-12$15.00/0

16872 • The Journal of Neuroscience, November 16, 2011 • 31(46):16872–16883

(Chang et al., 2003) as well as altered dendritic architecture (Bjork-blom et al., 2005). Second, the dopaminergic neurons of bothJnk2�/� and Jnk3�/� mice resist MPTP-induced cell death (Hunotet al., 2004). Thirdly, Jnk1/Jnk2 double mutant mice die at E11.5 withdefective neural tube morphogenesis and reduced apoptosis in thehindbrain but increased apoptosis and caspase activation in the fore-brain (Kuan et al., 1999; Sabapathy et al., 1999), showing that JNKsignaling has both pro- and anti-apoptotic effects on the developingbrain. Last, JNK signaling is involved in neuronal migration duringbrain development. Cortical neuronal migration was retarded byoverexpression of a dominant-negative form of JNK (Kawauchi et

al., 2003), and specific deletion of MKK4 inthe CNS caused a delay in neuronal radialmigration in the cerebral cortex as well asmisalignment of Purkinje cells in the cere-bellum (Wang et al., 2007). Together, thesereports demonstrate that the proper controlof JNK signaling is required for normalbrain development; however, the specificrole of MKK7 in this control has yet to bedefined.

Here we report the generation of anovel mouse model, Mkk7flox/flox Nestin-Cre mice, in which the murine mkk7 geneis specifically deleted in neural stem cellsand postmitotic neurons. We found thatMkk7flox/flox Nestin-Cre mice display phe-notypes different from those previouslyreported for Mkk4flox/flox Nestin-Cre mice,indicating that MKK7 has unique andcrucial functions in the developing brain.

Materials and MethodsAnimals. Mice carrying the mkk7 flox allelewere described previously (Schramek et al.,2011). Mkk7flox/flox mice were crossed to

Figure 1. The MKK7-JNK signaling pathway is activated in the developing brain and essential for postnatal viability. A, Analysis of JNK activity during mouse development. Extracts of brains fromWT mice at the indicated stages of development were prepared and immunoblotted to detect JNK activity. B, Analysis of MKK7 activity in WT adult mouse brain. MKK7 kinase activity was assayedas described in Materials and Methods. Testis, Negative control. C, CNS-specific targeting of the murine mkk7 gene. The genomic mkk7 locus, the mkk7 targeting vector, the predicted structure of themutated mkk7 locus, and the floxed and deleted mkk7 alleles are depicted. Black boxes, mkk7 exons; black arrowheads, loxP sites; white arrowheads, FRT sites; X, XbaI; H, HindIII; B, BamHI; P, PstI.The Neo and DTA cassettes and the probe used for Southern blotting are indicated. D, Confirmation of MKK7 deletion. HindIII-restricted genomic DNA was prepared from mice of the indicatedgenotypes at E16.5 and subjected to Southern blotting with the probe in C. E, Suppression of JNK activity by MKK7 disruption in the CNS. Left, Extracts of brains of E18.5 embryos of the indicatedgenotypes were immunoblotted to detect the indicated proteins. Right, Normalized protein expression levels for MKK7/Gapdh, phospho-JNK/JNK, and phospho-MKK4/MKK4 in the extracts in theleft panel were determined by densitometry and are shown in the histograms. F, Mkk7 disruption results in death immediately after caesarean section. Control and Mkk7flox/flox Nestin-Cre embryoswere obtained by caesarean section at E18.5 and subjected to gross examination. All control mice started breathing (13/13), but most Mkk7flox/flox Nestin-Cre mice (9/10) did not and diedimmediately. Scale bar, 5 mm.

Figure 2. Lack of Mkk7 causes abnormal brain development and reduced axon tracts in the cerebrum. A, Gross appearance of wholebrains of control and Mkk7flox/flox Nestin-Cre embryos at E18.5. Results are representative of embryos examined per genotype. Scale bar, 1mm. B–D, Histological analysis of brain defects in Mkk7flox/flox Nestin-Cre mice. Horizontal sections of the brains of control and Mkk7flox/flox

Nestin-Cre embryos at E18.5 were stained with cresyl violet. Left, Forebrains with inset boxes indicating CC (B), AC (C), and IC (D) are shown.Scale bar, 500 �m. Right, Inset boxes in left panels are magnified. Scale bar, 500 �m.

Table 1. MKK7 is critical for postnatal viability

Stage No. of litters

Number of embryos or newborns with genotype

Total Mkk7flox/� Mkk7flox/floxMkk7flox/�

Nestin-CreMkk7flox/flox

Nestin-Cre

E12.5 2 15 6 3 4 2E14.5 6 46 10 11 17 8E16.5 6 51 12 17 9 13E18.5 7 54 14 9 14 17P1 5 29 9 7 13 0

Mkk7flox/� Nestin-Cre mice were crossed with Mkk7flox/flox mice and the genotypes of embryos or neonates weredetermined by PCR at the indicated stages.

Yamasaki et al. • MKK7 Regulates Axon Elongation In Vivo J. Neurosci., November 16, 2011 • 31(46):16872–16883 • 16873

Nestin-Cre transgenic mice expressing Cre re-combinase under the control of the mouse nes-tin promoter and the second intronic neuralenhancer (Imai et al., 2006; Okada et al., 2006).The resulting control (Mkk7flox/�, Mkk7flox/flox

and Mkk7flox/� Nestin-Cre) and mutant(Mkk7flox/flox Nestin-Cre) mice were reared on anormal 12 h light/dark schedule. The day whena plug was observed in a female mouse wasdesignated as embryonic day 0.5 (E0.5), andthe day of birth was termed postnatal day 0(P0). Mouse genotypes were determined byPCR and Southern blotting. For all experi-ments, only littermate mice from the samebreeding were used. All procedures were per-formed in accordance with a protocol ap-proved by the Tokyo Medical and DentalUniversity Animal Care Committee.

In utero electroporation. In utero electropo-ration was performed as described previously(Saito, 2006; Tabata and Nakajima, 2008) withslight modifications. Briefly, pregnant micewere anesthetized with sodium pentobarbitaland the uterine horns were exposed. Approxi-mately 1–2 �l of DNA solution (1–2.5 mg/ml)was injected into the lateral ventricle of eachembryonic brain using a pulled glass micropi-pette. Each embryo within its uterus was thenplaced between tweezer-type electrodes with adiameter of 5 mm (CUY650-P5; NEPA Gene).Square electric pulses (45 V, 50 ms) werepassed five times at 1 s intervals using an elec-troporator (ECM830, BTX). Care was taken toquickly place embryos back into the abdominalcavity to avoid excessive temperature loss. Thewall and skin of the abdominal cavity were su-tured, and embryos were allowed to developnormally.

In vivo neuronal migration. Pregnant mice atE15.5 were injected intraperitoneally with bro-modeoxyuridine (BrdU) (50 mg/kg of bodyweight). Embryos were dissected and fixed atE18.5 and brains were processed for paraffinsectioning. Deparaffinized and rehydrated 6�m coronal sections were treated with 2N HClat 37°C for 30 min and neutralized with 0.1%boric acid buffer, pH 8.5. After blocking in 2% skim milk in PBS contain-ing 0.1% Triton X-100, sections were subjected to immunohistochemicalanalysis as described below.

Dissociation and culture of cortical neurons. At E16.5, cortices were dis-sected from Mkk7flox/flox Nestin-Cre, Mkk7flox/flox, Mkk7flox/�, or wild-typeembryos that had been electroporated at E15.5. The dissected corticeswere placed in culture and incubated for 20 min with papain solutionplus DNase I at 37°C, followed by mechanical dissociation in dissociationmedium (EBSS; Sigma-Aldrich). Cortical neurons were plated ontopoly-L-lysine-coated coverslips, and maintained in Basal Medium Eagle(Invitrogen) containing 1 mM L-glutamine (Sigma), HBSS (Invitrogen),25% (v/v) horse serum (Invitrogen), 6.6 mg/ml dextrose (Sigma), peni-cillin/streptomycin (Invitrogen), and 1 mM HEPES, pH 7.4 (Invitrogen).Cortical neuron cultures were maintained in 5% CO2 at 37°C.

Immunoblotting. Immunoblotting was performed as previously de-scribed (Ura et al., 2007) with slight modifications. Proteins were ex-tracted from tissues in TNE buffer [20 mM Tris-HCl, pH 7.4, 150 mM

NaCl, 1 mM EDTA, 1 mM EGTA, 0.5% Nonidet P40, 5% (w/v) glycerol,1 mM PMSF, 200 mM NaF, 200 �M Na3VO4, 10 �g/ml aprotinin]. Totalextracts were clarified by centrifugation (20,000 � g for 10 min at 4°C)and soluble proteins in the supernatants were quantified using the BCAProtein Assay Kit (PIERCE). Protein extracts were fractionated by SDS/PAGE and transferred to a PVDF membrane, which was incubated in

blocking solution (2% skim milk in TBS) for 1 h. The blocked membranewas incubated overnight in TBS containing 5% BSA plus antibodies rec-ognizing phospho-JNK (Cell Signaling Technology), JNK1 (46 kDa and55 kDa splicing variants; Santa Cruz Biotechnology), MKK7 (Cell Signal-ing Technology), GAPDH (Millipore Bioscience Research Reagents),phospho-MKK4 (Cell Signaling Technology), MKK4 (Santa Cruz Bio-technology), phospho-DCX (Gdalyahu et al., 2004), DCX (Gdalyahu etal., 2004), phospho-neurofilament heavy chain (Covance), phospho-MAP1B (Covance), MAP1B (Sigma), or tubulin (Invitrogen). The mem-brane was then washed in TBS/Tween 20 (0.05%), incubated for 1 hwith anti-mouse/rabbit horseradish peroxidase-conjugated antibod-ies (Jackson ImmunoResearch Laboratory), and washed three timesin TBS/Tween 20. Proteins were visualized using Immobilon-HRP(Millipore) or the SuperSignal West Femto Kit (Pierce) and Chemi-Doc XRS (Bio-Rad).

In vitro kinase assay. Assay of MKK7 activity was as described previ-ously (Nakagawa et al., 2010). Endogenous MKK7 proteins were immu-noprecipitated with anti-MKK7 (KN-004) monoclonal antibody (Wadaet al., 2001). Immunocomplexes were washed three times with lysis buf-fer and three times with kinase reaction buffer, which consisted of 10 mM

MgCl2, 50 mM Tris-HCl, pH 7.5, and 1 mM EGTA. MKK7 activity onbeads was measured using an in vitro MAPK kinase assay using 100 �M

ATP and GST-JNK1 as the substrate. Reactions were terminated after 30

Figure 3. Loss of MKK7 causes accumulations of filamentous structures and autophagic vacuoles. Brains of control and Mkk7flox/flox

Nestin-Cre embryos at E18.5 were analyzed by electron microscopy. The cortex (A) and striatum (B) are shown. Black arrows, Axons; blackarrowheads, mitochondria; white arrows, autophagic vacuoles; white arrowheads, swollen mitochondria; N, nuclei. Inset boxes, A, Accu-mulation of filamentous structures; B, top, swollen mitochondrion; B, bottom, autophagic vacuole. Scale bar, 2 �m.

16874 • J. Neurosci., November 16, 2011 • 31(46):16872–16883 Yamasaki et al. • MKK7 Regulates Axon Elongation In Vivo

min at 30°C by the addition of 4� SDS gel sample buffer. The reactionproduct was detected using anti-phospho-JNK antibody.

Nissl staining. Nissl staining was performed as described previously (Todaet al., 2008) with slight modifications. Brains were fixed at E18.5 and pro-cessed for paraffin sectioning. Deparaffinized and rehydrated 6 �m horizon-tal sections were subjected to Nissl staining with 0.2% cresyl violet.

Immunostaining. Immunohistochemistry was performed as describedpreviously (Kawasaki et al., 2000). For embryos, brains were isolated andfixed overnight in 4% paraformaldehyde. For postnatal mice, the animalswere deeply anesthetized with pentobarbital and transcardially perfusedwith 4% paraformaldehyde. Fixed tissues were cryoprotected by over-night immersion in 30% sucrose, followed by embedding in OCT com-pound. Sections of 14 �m thickness were attached to glass slides, whereassections of 50 �m thickness were floated on PBS. For cultures of disso-ciated neurons, cells were fixed by incubation in 4% paraformaldehydefor 5 min at 37°C.

For immunostaining, tissue sections and isolated cells were permeabil-ized with 0.1– 0.5% Triton X-100 in PBS and incubated overnight withprimary antibodies, including those recognizing phospho-JNK (Cell Sig-naling Technology), TAG-1 (DSHB), L1 (Millipore Bioscience ResearchReagents), cleaved caspase-3 (PharMingen), �-tubulin III (Covance),Brn2 (Santa Cruz Biotechnology), Satb2 (Abcam), Ctip2 (Abcam),Foxp2 (Abcam), Tbr2 (Abcam), or GFP or RFP (both from MBL). Im-munostained tissues and cells were then incubated with Alexa488-and/or Cy3-conjugated secondary antibodies and 1 �g/ml Hoechst33342, followed by washing and mounting. Epifluorescent microscopywas performed using an Axioimager A1 microscope (Carl Zeiss) andBZ9000 (Keyence). Confocal microscopy was performed using anLSM510 microscope (Carl Zeiss).

Electron microscopy. Tissues were fixed by immersion first in 1.5%paraformaldehyde plus 3% glutaraldehyde in 0.1 M phosphate buffer,pH 7.3, and then in an aqueous solution of 1% OsO4. Fixed samples were

Figure 4. Axonal defects in the cortex and corpus callosum of Mkk7flox/flox Nestin-Cre mice. A, Suppression of JNK phosphorylation in axons. Coronal sections of the cortex in control and Mkk7flox/flox

Nestin-Cre embryos at E18.5 were immunostained with anti-phospho-JNK and anti-L1 antibodies. DAPI staining was used to visualize nuclei. CP, Cortical plate; IZ, intermediate zone. B, C, Coronalsections of the cortex (B) and the CC (C) in control and Mkk7flox/flox Nestin-Cre embryos at E18.5 were immunostained with anti-TAG-1 and anti-L1 antibodies as for A. Scale bar: A–C, 200 �m.


Figure 5. Loss of MKK7 does not influence apoptosis or neuronal differentiation in the brain. A, B, Coronal sections of cortex (A) and CC (B) in control and Mkk7flox/flox Nestin-Cre embryos at E18.5were immunostained with anti-cleaved caspase-3 and anti-�-tubulin III antibodies. C, D, Coronal sections of the cortex in control and Mkk7flox/flox Nestin-Cre embryos at E16.5 (C) or E18.5 (D) wereimmunostained with antibodies recognizing the neuronal markers Brn2, Satb2, Ctip2, Foxp2, or Tbr2. Scale bar, 200 �m.


embedded in Epon812, and thin sections (70 – 80 nm) were cut andstained with uranyl acetate and lead citrate for observation under a Jeol-1010 electron microscope (Jeol) at 80 kV.

Plasmids. All genes were expressed under the control of the CAG pro-moter. pCAG-NLS-Cre and pCAG-floxed-polyA-EGFP were kindly pro-vided by Dr. Fumio Matsuzaki (RIKEN-CDB, Japan). pCAG-EGFP-MKK7, pCAG-EGFP-MKK4, pCAG-MKK7-JNK, pCAG-c-Jun-IRES-EGFP, and pCAG-EGFP-JBD were constructed in the Nishina laboratoryusing standard protocols. Plasmids were purified using the Endofree Plas-mid Maxi kit (Qiagen).

Statistical analysis. Data were analyzed using the paired Student’s t test fortwo-tailed distributions. Significance levels were as follows: *p � 0.05, **p �0.01, ***p � 0.001, ****p � 0.0001.

ResultsMkk7 flox/flox Nestin-Cre mice die at P0 without breathingBecause MKK7 total knock-out mice are embryonic lethal, it hasnot been possible to examine MKK7 functions in the brain. Toconfirm that MKK7-JNK signaling occurs during normal braindevelopment, we established a time course of JNK phosphoryla-tion and MKK7 activity in wild-type (WT) mouse brain. Extractsof brains of embryos from E12.5 to adult were prepared and

immunoblotted to detect total and acti-vated JNK. MKK7 activity was measured byin vitro kinase assay. We found that JNK andMKK7 were highly activated from the E16.5to the adult stages (Fig. 1A,B, and data notshown). These data establish that MKK7-JNK signaling occurs during normal braindevelopment.

To investigate the role of this MKK7-JNK signaling in vivo, we made a mutantmouse carrying a mkk7 flox allele (Fig.1C), and crossed it with a transgenicmouse line that expresses Cre under thecontrol of the nestin promoter (Imai et al.,2006). We confirmed the deletion of mkk7in Mkk7flox/flox Nestin-Cre brain at E16.5by Southern blotting (Fig. 1D). As ex-pected, levels of MKK7 and phosphory-lated JNK proteins were markedly reducedin the brains of these mice, but, unexpect-edly, phospho-MKK4 was upregulated (Fig.1E). These results suggest that, in the ab-sence of one JNK regulator (MKK7), the ac-tivation of another JNK regulator (MKK4)may increase in compensation.

To investigate whether MKK7 was im-portant for postnatal viability, we inter-crossed Mkk7flox/� Nestin-Cre andMkk7flox/flox mice and determined the ge-notypes of embryos and postnatal pups.Surprisingly, although Mkk7floxflox Nestin-Cre mice appeared normal through allembryonic stages, no mutants were de-tected on P1 (Table 1). To determinewhether Mkk7flox/flox Nestin-Cre mice diedat birth, we performed caesarean sectionsand found that Mkk7flox/flox Nestin-Cremice were indistinguishable from con-trols at E18.5. However, all but oneMkk7flox/flox Nestin-Cre mouse died im-mediately without breathing after caesar-ean section (Fig. 1F). MKK7 is thereforevital for postnatal survival.

Abnormal brain development in Mkk7flox/flox Nestin-Cre miceThe death of Mkk7flox/flox Nestin-Cre mutants at birth suggestedthat there might be defects in the brains of these animals. Toassess brain histology, sections of various brain structures wereprepared from control (Mkk7flox/�, Mkk7flox/flox, and Mkk7flox/�

Nestin-Cre) and Mkk7flox/flox Nestin-Cre mice at E18.5 and stainedwith cresyl violet. Although Mkk7flox/flox Nestin-Cre whole brainwas grossly indistinguishable from control littermate whole brain(Fig. 2A), histological analysis showed that Mkk7flox/flox Nestin-Cre brain had larger ventricles and a reduced striatum comparedto control brain (Fig. 2B). In addition, whereas the corpus callo-sum (CC) and the anterior commissure (AC) were clearly formedin control E18.5 brain, these commissural axon tracts were barelydetectable in E18.5 Mkk7flox/flox Nestin-Cre brain (Fig. 2B,C).Moreover, although there was no defect in the mutant hippocam-pal commissure (data not shown), the internal capsule (IC),which consists of axonal fibers connecting the cerebral cortex andsubcortical brain regions, was reduced in size in Mkk7flox/flox

Figure 6. MKK7 deletion delays neuronal radial migration in the cortex. A, Control and Mkk7flox/flox Nestin-Cre embryos werelabeled in utero with BrdU at E15.5 and coronal sections were prepared at E18.5 and analyzed by immunostaining with anti-BrdUantibody. White, BrdU. Scale bar, 100 �m. CP, cortical plate; IZ, intermediate zone. B, Distribution of BrdU-positive nuclei. Theimmunostained cortices in A were divided into four equal bins from the pial side to the ventricular side. The percentage of the totalnumber of immunostained nuclei in a particular bin was determined and the results plotted in a histogram. Data shown are themean � SD (n � 6). *p � 0.05; **p � 0.01.


Nestin-Cre brain (Fig. 2D). These data indicate that MKK7 reg-ulates the formation of axon tracts during brain development.

To further examine the effect of MKK7 deletion on axontracts, we used electron microscopy to look for ultrastructuralalterations in Mkk7flox/flox Nestin-Cre brain. We found that abnor-mally electron-dense axons were present in Mkk7flox/flox Nestin-Cre cortex and striatum (Fig. 3A,B), and that these structurescontained an accumulation of intermediate filaments (Fig. 3A,inset). In addition, autophagic vacuoles and swollen mitochon-dria were detected in the striatum, cortex, and brainstem ofMkk7flox/flox Nestin-Cre brain (Fig. 3B and data not shown). Thus,MKK7-mediated regulation of JNK may be important for sup-pressing autophagy and controlling the distribution of interme-diate filaments in the developing brain.

Impaired axon formationOur histological analyses led us to hypothesize that MKK7-JNK sig-naling might normally be activated in the axons of WT developingbrain. As expected, phospho-JNK coimmunostained with L1, whichis a cell adhesion molecule expressed on axons, in control brain butwas reduced in Mkk7flox/flox Nestin-Cre brain (Fig. 4A). We thenperformed a more detailed analysis by immunostaining to detectboth L1 and transient axonal glycoprotein-1 (TAG-1), which is ex-pressed in corticofugal axons (Wolfer et al., 1994; Denaxa et al.,2001). In Mkk7flox/flox Nestin-Cre brain at E18.5, we found that num-bers of L1-positive as well as TAG-1-positive axons were decreased inthe cortex and CC compared to controls (Fig. 4B,C). These resultsindicate that MKK7 has a particularly important role in the forma-tion of TAG-1-positive axons.

Because previous reports implicated JNK in regulating apoptosisduringbraindevelopment(Kuanetal., 1999;Sabapathyetal., 1999),weexamined whether apoptosis was altered in Mkk7flox/flox Nestin-Crebrain. Immunostaining with antibody specific for cleaved caspase-3 re-vealed that, like control brain, there were few apoptotic cells in theMkk7flox/flox Nestin-Cre cortex (Fig. 5A). Apoptotic cells were also rare inthestriatumofbothMkk7flox/flox Nestin-Creandcontrolbrains(datanotshown). A few apoptotic cells were observed surrounding the mutantCC but were no more frequent than in control brains (Fig. 5B). Thus, alack of MKK7 does not impair the regulation of apoptosis in the devel-oping brain.

We next investigated whether loss of MKK7 caused any alter-ations to neuronal differentiation. Cortical neurons can be iden-tified by immunostaining with antibodies recognizing neuronalmarkers such as Brn2, Satb2, Ctip2, Foxp2, and Tbr2 (Bulfone etal., 1999; McEvilly et al., 2002; Ferland et al., 2003; Arlotta et al.,2005; Britanova et al., 2005). We detected expression of all ofthese markers in Mkk7flox/floxNestin-Cre cortex (Fig. 5C,D), indi-cating that cortical neurons differentiate normally withoutMKK7. Next, because it was reported that the major JNK substratec-Jun is involved in the specification of GABAergic/glutamatergicneurons in the spinal cord of Xenopus tropicalis (Marek et al., 2010),we performed an in situ hybridization analysis of Mkk7flox/flox Nestin-Cre spinal cord using probes for gad1 and vglut2, which are expressedin GABAergic or glutamatergic neurons, respectively. However, nosignificant alterations to expression levels of gad1 or vglut2 weredetected between control and Mkk7flox/flox Nestin-Cre mice (data notshown). These data show that, as was true for cortical neurons, thedifferentiation of GABAergic and glutamatergic neurons is not af-fected by MKK7 deletion.

Delayed neuronal radial migrationPrevious reports showed that overexpression of dominant-negative JNK or MKK4 disruption can delay neuronal radial mi-

gration in the cerebral cortex (Kawauchi et al., 2003; Wang et al.,2007). To determine whether MKK7 also affects this process, weperformed a timed pregnancy study of BrdU incorporation in thedeveloping cerebral cortex. Embryos were labeled in utero withBrdU at E15.5 and brain sections prepared at E18.5 were immu-nostained with anti-BrdU antibody. We found that, in Mkk7flox/flox

Nestin-Cre brain, BrdU-positive cells were dispersed throughoutthe ventricular zone (VZ), intermediate zone, and cortical plate,with a small proportion of BrdU-positive cells located in Bin1. Incontrol brain, a significantly larger proportion of BrdU-positivecells was located in Bin1 (Fig. 6A,B). Thus, like MKK4, MKK7influences the radial migration of neurons in the cortex.

Figure 7. Deletion of mkk7 in layer 2/3 neurons prevents axon elongation in vivo. A–C, Layer-specific MKK7 deletion. pCAG-NLS-Cre and pCAG-floxed-polyA-EGFP were introduced by in utero elec-troporation into the VZ of control Mkk7flox/� and Mkk7flox/flox embryos at E15.5. Brains were fixed onthe indicated postnatal days and coronal sections of the cortex were immunostained with anti-GFPantibody.Hp.,Hippocampus.Scalebar,1mm. D,Higher-magnificationviewofaxonsinthecorticesofthe Mkk7flox/� and Mkk7flox/flox mice in C. Scale bar, 200 �m. E, Higher-magnification view of theinset boxes in D, focusing on the white matter (broken lines). Scale bar, 100 �m.


MKK7 deletion in layer 2/3 neurons results in thedisappearance of contralateral projecting axons in vivoOur immunostaining experiments indicated that axon tractswere abnormal in the absence of MKK7. However, axon elonga-tion is regulated not only by the neuron itself but also by envi-

ronmental factors such as nerve growth factors. To determinewhether the axon defect, we observed was due to the effects ofMKK7 deletion on neuronal axon elongation or on environmen-tal factors, we generated mice in which MKK7 disruption wasinduced specifically in layer 2/3 neurons. We used in utero elec-

Figure 8. Deletion of mkk7 in layer 2/3 neurons reduces neurite length in vitro. A–C, Reduced axon elongation. The VZ of control (n � 6) and Mkk7flox/flox Nestin-Cre (n � 3) embryos at E15.5 waselectroporated with pCAG-EGFP. Brains were dissected at E16.5 and primary neuron cultures were established. A, After 2, 4, or 6 DIV, neurons were fixed and stained with anti-GFP antibody (upperpanels; scale bar, 100 �m). B, The lengths of the longest neurites of individual neurons in the cultures in A were measured at 2, 4, and 6 DIV (�100 neurons examined in each 2 DIV and 4 DIV culture,and �40 neurons in each 6 DIV culture). The results are plotted as a histogram and are the mean � SE for each time point (Mkk7flox/flox Nestin-Cre, n � 3; control, n � 6). C, The distribution of thelongest neurite lengths was plotted for each set of cultures in A. Results are the mean � SD (Mkk7flox/flox Nestin-Cre, n � 3; control, n � 6). D, Cell-autonomous defect. The VZ of control andMkk7flox/flox embryos at E15.5 was electroporated with pCAG-NLS-Cre plus pCAG-floxed-polyA-EGFP, and primary neuron cultures were established at E16.5. After 4 DIV, neurons were fixed andstained with anti-GFP-antibody (upper; scale bar, 50 �m). The lengths of the longest neurites of individual neurons were measured, and the distribution of these lengths was plotted (lower). Results are themean�SD(n�3).E,EffectofJNKinhibition.TheVZofWTembryosatE15.5waselectroporatedwithpCAG-EGFPorpCAG-EGFP-JBD,andprimaryneuroncultureswereestablishedatE16.5.After2DIV,neuronswere fixed and stained (left; scale bar, 50 �m). The distribution of neurite lengths was analyzed as for E (right). Results are the mean � SD (n � 3). For A–D, *p � 0.05, ***p � 0.001, ****p � 0.0001.


troporation to introduce pCAG-NLS-Cre and its reporter plas-mid pCAG-floxed-polyA-EGFP into the VZ of control andMkk7flox/flox embryos at E15.5. We then killed the pups at P4, P10,or P31 and examined them for GFP fluorescence (representingsuccessful Cre recombination). GFP-positive neurons were de-tected in layer 2/3 neurons in both Mkk7flox/� and Mkk7flox/flox

brains through all postnatal stages examined. However, unlikecontrols, GFP-positive neurons in Mkk7flox/flox brain were notable to elongate axons to reach the contralateral cortex by P4 (Fig.7A), nor was contralateral elongation detected at P10 or P31 inthe mutant brains (Fig. 7B,C). Interestingly, when we investi-gated the morphology of these faulty axons in detail at P31, wefound that both control and Mkk7flox/flox layer 2/3 neurons wereable to elongate axons radially in the cerebral cortex such thattheir branches projected to layer 5 (Fig. 7D). However, in themutant brain, axons running tangentially in the white matter haddisappeared (Fig. 7E). When we attempted to determine whetherthe disappearance of these contralaterally projecting axons wasdue to a failure in axon guidance, we found no evidence of anaxon guidance defect (data not shown). These results show in vivothat MKK7 regulates axon elongation in a cell-autonomous man-ner, and that MKK7 is required for the contralateral projection ofaxons of layer 2/3 neurons in the white matter.

We also examined axon elongation of layer 2/3 neurons in vitro.We introduced GFP into control and Mkk7flox/flox Nestin-Cre brainsby in utero electroporation at E15.5 and prepared cultures of disso-ciated neurons at E16.5. We then measured the lengths of the longestneurites appearing in the cultures and found that there were moreneurons with shorter neurites in Mkk7flox/flox Nestin-Cre cultures af-ter 2, 4, or 6 d in vitro culture (DIV) compared with control cultures(Fig. 8A–C). To determine whether this defect in axon growth wasinduced a cell-autonomous manner, we disrupted MKK7 inMkk7flox/flox layer 2/3 neurons using in utero electroporation of NLS-Cre and floxed-polyA-EGFP, and cultured control and mutantGFP-positive neurons for 4 d. Upon measurement of the longestneurites appearing in these cultures, we found that there were moreneurons with shorter neurites in Mkk7flox/flox cultures at 4 DIV com-pared with Mkk7flox/� cultures (Fig. 8D). These results confirm invitro that MKK7 regulates axon elongation in a cell-autonomousmanner.

To determine whether our observations were due to effects onJNK signaling, we investigated whether inhibition of JNK causedthe same neuronal phenotype as MKK7 disruption. WT layer 2/3neurons were prepared after in utero electroporation of a plasmidexpressing the JNK binding domain of JIP1 (JBD), which acts asa JNK inhibitory peptide (Westerlund et al., 2011). We foundthat, like MKK7 deletion, JNK inhibition resulted in more neu-rons with shorter axons (Fig. 8E). These results support our hy-pothesis that MKK7’s effects on axon elongation are likely due toMKK7’s regulation of its downstream target JNK.

Reduced phosphorylation of cytoskeletal componentsWe next determined whether the loss of MKK7 in neurons alteredthe phosphorylation of JNK substrates. Because phosphorylation ofc-Jun was reduced in Mkk7flox/flox Nestin-Cre neurons (Fig. 9A), weperformed rescue experiments in which plasmids expressing MKK7,MKK4, MKK7-JNK fusion protein (constitutively active JNK), orc-Jun were coelectroporated along with NLS-Cre and floxed-polyA-mCherry or -GFP into the VZ of E15.5 control and mutant embryos.Coelectroporation of MKK7 (Fig. 9B) and constitutively active JNK(Fig. 9D) allowed mutant neurons to recover their ability to con-tralaterally elongate axons, but the same was not true for coelectro-poration of either MKK4 (Fig. 9C) or c-Jun (Fig. 9E). These data

suggest that MKK7 regulates axon elongation through JNK, butthat this signaling pathway does not involve transcription medi-ated by c-Jun.

In Mkk4flox/flox Nestin-Cre mice, the phosphorylation of cer-tain neurofilament and MAPs is reportedly altered (Wang et al.,2007). Neurofilament heavy chain (NF-H) is a neuron-specificintermediate filament whose C-terminal tail is constitutively phos-phorylated in axons by several proline-directed kinases, includingJNK. When we investigated phosphorylation levels of NF-H inMkk7flox/flox Nestin-Cre brain, we found that NF-H phosphorylationwas reduced (Fig. 10A). The axon elongation and radial migrationrequired for normal brain development also rely on MAP-dependent regulation of microtubule dynamics, and MAP activity iscontrolled by phosphorylation mediated by a number of proteinkinases, including JNK. When we tested whether MAP phosphory-lation was altered in Mkk7flox/flox Nestin-Cre brain, we found thatphosphorylation levels of MAP1B were reduced (Fig. 10B). In addi-tion, the phosphorylation of doublecortin (DCX) was suppressedby loss of MKK7 (Fig. 10C), an alteration that does not occur afterdisruption of MKK4 (Wang et al., 2007). These data show that,within the developing brain, MKK7 regulates the phosphoryla-

Figure 9. Contralateral axon projection is rescued by electroporation of MKK7 and active-JNK butnot by MKK4 or c-Jun. A, Decreased c-Jun phosphorylation. Brain extracts prepared from control andMkk7flox/flox Nestin-Cre embryos at E18.5 were immunoblotted to detect total c-Jun and phospho-c-Jun. B–E, Reversal of effects of MKK7 disruption. pCAG-NLS-Cre and pCAG-floxed-polyA-mCherry or-GFP were coelectroporated along with MKK7 (0.25 mg/ml) (B), MKK4 (0.25 mg/ml) (C), MKK7-JNK(0.05 mg/ml) (D), or c-Jun (1.0 mg/ml) (E) into the VZ of control and Mkk7flox/flox embryos at E15.5.Brains were fixed at P10 and coronal sections of cortices were immunostained with anti-RFP or GFPantibody. Arrows indicate axons rescued by coelectroporation. Hp, Hippocampus. Scale bar, 1 mm.


tion of not only NF-H and MAP1B, which are also regulated byMKK4, but also DCX.

We next examined whether microtubule stabilization couldameliorate the axonal defects caused by MKK7 disruption. Culturesof Mkk7flox/flox Nestin-Cre primary neurons were treated with Taxol,a drug that stabilizes microtubules and induces axon formation (Hi-rai et al., 2011). We found that Taxol treatment did indeed increasethe lengths of axons in Mkk7flox/flox Nestin-Cre brain, but that thesestructures were still shorter than axons in Taxol-treated control cul-tures (data not shown). Finally, we examined microtubule structuresin sections of axons directly by electron microscopy and found thatMkk7flox/flox Nestin-Cre brain contained axons with fewer andshorter microtubules than control brain (Fig. 10D). Thus, the regu-latory effects of MKK7 on molecules critical for microtubule struc-tures involve more than mere microtubule stabilization and do notoverlap the effects of MKK4.

DiscussionRole of MKK7 in axon elongationIn this study, we investigated the physiological functions ofMKK7 in the developing brain. Our most striking finding isthat MKK7 regulates the elongation of axons by layer 2/3 neu-rons. Mkk7flox/flox Nestin-Cre brains displayed defective forma-tion of axon tracts and decreased numbers of TAG-1-positiveaxons (Figs. 2, 4), a phenotype never observed in JNK-deficientmice. Jnk1/2 double mutant mice die during early embryogenesis,precluding analysis of the developing brain. No brain defectshave been reported for Jnk2 and/or Jnk3 knock-out mice (Hunotet al., 2004), and although Jnk1�/� mice display abnormalities ofaxon maintenance and radial migration, axon formation is nor-mal in these mutants (Chang et al., 2003; Westerlund et al., 2011).These results indicate that there is redundancy among JNK1, 2,and 3 functions that obscures the effects of JNK signaling on axonelongation because all of these kinases are expressed in the brain.Our report on mice lacking the JNK activator MKK7 specificallyin neural stem cells and postmitotic neurons is the first to dem-onstrate that JNK signaling regulates axon elongation in vivo.

By using in utero electroporation, weshowed that MKK7 controls axon elonga-tion in a cell-autonomous manner (Figs. 7,8). Our work has also uncovered a very in-teresting phenotype affecting neuronal con-tralateral axon projection. It is well knownthat axons of layer 2/3 neurons run radiallyin the ipsilateral cortex to the white matter,and then run tangentially and project to thecontralateral cortex. In the ipsilateral side,layer 2/3 axons normally extend branchesprojecting to layer 5 and form barrel nets,which are axonal trajectories made afterthe initial appearance of barrels (Seharaet al., 2010). In our study, layer 2/3 neu-rons lacking MKK7 did not extend ax-ons running tangentially in the whitematter or form contralateral projec-tions, although the axons of these neu-rons did run radially in the ipsilateralcortex and formed branches and barrelnets (Fig. 7 D, E). Our results suggestthat axon elongation is regulated by dif-ferent mechanisms in the cerebral cor-tex and in the white matter, and thatMKK7 plays an important role only inthe latter case.

Role of MKK7 in MAP regulationProper regulation of microtubule dynamics is critical for axonelongation and radial migration, and this control is mediatedmainly by MAPs such as MAP1B and DCX that bind to microtu-bules and stabilize them. Double mutant mice such as the Map1b/Tau or Dcx/doublecortin-like kinase (Dclk) strains display axonalmalformation during brain development (Takei et al., 2000; Koi-zumi et al., 2006). Importantly, the ability of MAPs to regulatemicrotubule dynamics depends on their phosphorylation by sev-eral protein kinases, including JNK (Kawauchi et al., 2003; Gda-lyahu et al., 2004). Phosphorylation of a MAP protein induces itsdisassociation from microtubules, which causes the microtubulesto enter a dynamic state. In our study, we found that the phos-phorylation levels of both MAP1B and DCX were reduced (Fig.10B,C), and that microtubule structures appeared abnormal inMkk7flox/flox Nestin-Cre axons (Fig. 10D). Based on our resultsand the literature, we propose that MKK7-JNK signaling controlsthe phosphorylation of MAP1B and DCX that regulates micro-tubule dynamics, and that this phosphorylation may be one ofseveral mechanisms promoting axon elongation.

Role of MKK7 in apoptosisIt has been previously reported that Jnk1/Jnk2 double mutantmice are embryonic lethal at E11.5 due to severely dysregulatedapoptosis in the brain (Kuan et al., 1999; Sabapathy et al., 1999).The results of these studies indicated that JNK regulates apoptosispositively in the forebrain and negatively in the hindbrain duringearly brain development. However, when we examined the regu-lation of apoptosis in our Mkk7flox/flox Nestin-Cre mice, we foundno abnormalities (Fig. 5), in line with previous analyses ofMkk4flox/flox Nestin-Cre mice (Wang et al., 2007). One explana-tion for this discrepancy is that JNK signaling may regulate apo-ptosis during early brain development but is not involved in suchapoptosis at later stages. Another possibility is that, becauseMKK4 remains intact in MKK7flox/flox Nestin-Cre brain, weak JNK

Figure 10. MKK7 regulates MAP phosphorylation. A–C, Brain extracts from control and Mkk7flox/flox Nestin-Cre embryos atE18.5 were immunoblotted to detect phospho-NF-H and tubulin (loading control) (A); phospho-MAP1B and total MAP1B (B); andphospho-DCX and total DCX (C). D, Defective microtubule structures in axons of Mkk7flox/flox Nestin-Cre brain. Electron micrographsof horizontal sections (upper) and cross-sections (lower) of axons in control and Mkk7flox/floxNestin-Cre brains are shown. Blackarrowheads, Microtubules. Scale bars, 200 nm.


activation could have occurred that was sufficient for normalregulation of apoptosis in the developing brain. We are currentlygenerating double conditional knock-out Mkk7flox/flox/Mkk4flox/flox

Nestin-Cre mice to resolve this issue.

Role of MKK7 in neuronal differentiationWhen we investigated whether MKK7 was involved in neuronaldifferentiation, all markers examined were expressed with essen-tially normal distributions (Fig. 5). These results indicate thatneuronal differentiation is not affected by MKK7 disruption inthe cortex. We also determined whether MKK7 was involved inregulating the activity of c-Jun needed for neurotransmitter spec-ification in the dorsal spinal cord. A recent study of Xenopustropicalis demonstrated that c-Jun regulates neurotransmitterspecification through transcriptional regulation of Tlx3 in thedorsal spinal cord of this animal (Marek et al., 2010). This groupshowed that c-Jun activation induces the formation of GABAer-gic neurons and suppresses glutamatergic differentiation. Be-cause c-Jun is a major substrate of JNK, we hypothesized that wemight see a reduction of GABAergic neurons and an increase inglutamatergic neurons in the spinal cords of Mkk7flox/flox Nestin-Cre mice due to suppression of c-Jun activity caused by the lack ofMKK7. However, our in situ hybridization analysis of gad1 andvglut2 expression showed no alterations in the dorsal spinal cordsof Mkk7flox/flox Nestin-Cre mice (data not shown). Thus, neu-rotransmitter specification is not influenced by MKK7 deletion.It remains possible that low levels of c-Jun phosphorylation stilloccurred in Mkk7flox/flox Nestin-Cre mice that were sufficient fornormal neurotransmitter specification. Alternatively, GABAer-gic/glutamatergic differentiation may be regulated by differentmechanisms in Xenopus tropicalis and Mus musculus.

Ultrastructural alterations in Mkk7 flox/flox Nestin-Cre miceElectron microscopy revealed abnormal accumulations of inter-mediate filaments in the axons of Mkk7flox/flox Nestin-Cre brain(Fig. 3A), as well as reduced phosphorylation of NF-H (Fig. 10A).NF-H is a neuron-specific intermediate filament, and it has beenreported that the C-terminal tail of NF-H regulates interactionsamong NFs and influences their axonal transport (Sasaki et al.,2009; Lee et al., 2011). It is therefore possible that JNK-mediatedphosphorylation of NF-H is required for normal NF-H distribu-tion, and that failure of this regulation due to loss of MKK7induces the accumulation of filamentous structures within ax-ons. Our electron microscopy analyses also showed that Mkk7flox/flox

Nestin-Cre brain contained numerous autophagic vacuoles (Fig.3B). Although it is possible that autophagy is a secondary effectinduced by the abnormal accumulation of intermediate filamentsin our mutants, it has been reported that JNK signaling is directlyinvolved in the formation of autophagic vacuoles and autophagiccell death (Byun et al., 2009; Shimizu et al., 2010; Kim et al., 2011;Xu et al., 2011). We are currently studying our Mkk7flox/flox

Nestin-Cre mice to determine whether MKK7-JNK signaling actsto suppress autophagy in the developing brain.

Comparison of MKK4 and MKK7 in the developing brainMKK7 and MKK4 are both JNK activators, and complete activa-tion of JNK requires the activities of both of these kinases(Kishimoto et al., 2003). In our Mkk7flox/flox Nestin-Cre mice,phosphorylated JNK was markedly reduced despite an upregula-tion of MKK4 (Fig. 1E). This result is consistent with that ofWang and colleagues, who showed that MKK4 deletion decreasesJNK activation to 1/5 of control values (Wang et al., 2007). To-gether, these findings suggest that both MKK7 and MKK4 are

required for optimal JNK activation in the developing brain. Fur-thermore, this optimal JNK activation is essential for normalneuronal radial migration in the cortex, since both Mkk7flox/flox

Nestin-Cre mice (Fig. 6) and Mkk4flox/flox Nestin-Cre mice (Wanget al., 2007) show defects in this process. However, other pheno-types do not overlap between MKK7flox/flox Nestin-Cre andMkk4flox/flox Nestin-Cre mice. Mkk7flox/flox Nestin-Cre mice die atbirth (Table 1, Fig. 1F), whereas Mkk4flox/flox Nestin-Cre micesurvive until age 3 weeks. Mkk7flox/flox Nestin-Cre mice (E18.5)display enlarged brain ventricles, diminished striatum, decreasedforebrain axon tracts, and reduced corticofugal axons (Figs. 2, 4), butnone of these defects has been observed (or at least has not beenreported) for Mkk4flox/flox Nestin-Cre mice. Thus, MKK7 has uniquefunctions in the developing brain that differ from those of MKK4.

Differences between MKK7 and MKK4 functions also appearat the molecular level. In Mkk4flox/flox Nestin-Cre brain, phos-phorylation levels of MAP1B are reduced, but DCX phosphory-lation is not altered (Wang et al., 2007). In contrast, we found thatthe phosphorylation levels of both MAP1B and DCX were re-duced in Mkk7flox/flox Nestin-Cre brain (Fig. 10B,C), suggesting adifference in substrates affected by MKK7-JNK versus MKK4-JNK signaling. In line with this hypothesis, the scaffold proteinJIP1 binds to JNK, MKK7, and DCX but not to MKK4 (Whit-marsh et al., 1998; Wang et al., 2007). We therefore propose thatdifferences in scaffold proteins and/or substrates involved inMKK7-JNK versus MKK4-JNK signaling pathways could causethe phenotypic divergence observed between Mkk7flox/flox Nestin-Cre and Mkk4flox/flox Nestin-Cre mice.

In conclusion, our results reveal an important and unique role forMKK7 during axon elongation in the developing brain, and rein-force the complex nature of JNK signaling and regulation in vivo.

ReferencesArlotta P, Molyneaux BJ, Chen J, Inoue J, Kominami R, Macklis JD (2005)

Neuronal subtype-specific genes that control corticospinal motor neurondevelopment in vivo. Neuron 45:207–221.

Asaoka Y, Nishina H (2010) Diverse physiological functions of MKK4 andMKK7 during early embryogenesis. J Biochem 148:393– 401.

Bjorkblom B, Ostman N, Hongisto V, Komarovski V, Filen JJ, Nyman TA,Kallunki T, Courtney MJ, Coffey ET (2005) Constitutively active cyto-plasmic c-Jun N-terminal kinase 1 is a dominant regulator of dendriticarchitecture: role of microtubule-associated protein 2 as an effector.J Neurosci 25:6350 – 6361.

Britanova O, Akopov S, Lukyanov S, Gruss P, Tarabykin V (2005) Novel tran-scription factor Satb2 interacts with matrix attachment region DNA elementsin a tissue-specific manner and demonstrates cell-type-dependent expressionin the developing mouse CNS. Eur J Neurosci 21:658–668.

Bulfone A, Martinez S, Marigo V, Campanella M, Basile A, Quaderi N, Gat-tuso C, Rubenstein JL, Ballabio A (1999) Expression pattern of the Tbr2(Eomesodermin) gene during mouse and chick brain development. MechDev 84:133–138.

Byun JY, Yoon CH, An S, Park IC, Kang CM, Kim MJ, Lee SJ (2009) The Rac1/MKK7/JNK pathway signals upregulation of Atg5 and subsequent autophagiccell death in response to oncogenic Ras. Carcinogenesis 30:1880–1888.

Chang L, Karin M (2001) Mammalian MAP kinase signalling cascades. Na-ture 410:37– 40.

Chang L, Jones Y, Ellisman MH, Goldstein LSB, Karin M (2003) JNK1 isrequired for maintenance of neuronal microtubules and controls phos-phorylation of microtubule-associated proteins. Dev Cell 4:521–533.

Davis RJ (2000) Signal transduction by the JNK group of MAP kinases. Cell103:239 –252.

Denaxa M, Chan CH, Schachner M, Parnavelas JG, Karagogeos D (2001)The adhesion molecule TAG-1 mediates the migration of cortical in-terneurons from the ganglionic eminence along the corticofugal fibersystem. Development 128:4635– 4644.

Enslen H, Raingeaud J, Davis RJ (1998) Selective activation of p38 mitogen-activated protein (MAP) kinase isoforms by the MAP kinase kinasesMKK3 and MKK6. J Biol Chem 273:1741–1748.


Ferland RJ, Cherry TJ, Preware PO, Morrisey EE, Walsh CA (2003) Charac-terization of Foxp2 and Foxp1 mRNA and protein in the developing andmature brain. J Comp Neurol 460:266 –279.

Gdalyahu A, Ghosh I, Levy T, Sapir T, Sapoznik S, Fishler Y, Azoulai D, Reiner O(2004) DCX, a new mediator of the JNK pathway. EMBO J 23:823–832.

Hibi M, Lin A, Smeal T, Minden A, Karin M (1993) Identification of anoncoprotein- and UV-responsive protein kinase that binds and potenti-ates the c-Jun activation domain. Genes Dev 7:2135–2148.

Hirai S, Banba Y, Satake T, Ohno S (2011) Axon formation in neocorticalneurons depends on stage-specific regulation of microtubule stability bythe dual leucine zipper kinase-c-Jun N-terminal kinase pathway. J Neu-rosci 31:6468 – 6480.

Hunot S, Vila M, Teismann P, Davis RJ, Hirsch EC, Przedborski S, Rakic P,Flavell RA (2004) JNK-mediated induction of cyclooxygenase 2 is re-quired for neurodegeneration in a mouse model of Parkinson’s disease.Proc Natl Acad Sci U S A 101:665– 670.

Imai F, Hirai S, Akimoto K, Koyama H, Miyata T, Ogawa M, Noguchi S,Sasaoka T, Noda T, Ohno S (2006) Inactivation of aPKClambda resultsin the loss of adherens junctions in neuroepithelial cells without affectingneurogenesis in mouse neocortex. Development 133:1735–1744.

Ito M, Yoshioka K, Akechi M, Yamashita S, Takamatsu N, Sugiyama K, HibiM, Nakabeppu Y, Shiba T, Yamamoto KI (1999) JSAP1, a novel junN-terminal protein kinase (JNK)-binding protein that functions as a scaf-fold factor in the JNK signaling pathway. Mol Cell Biol 19:7539 –7548.

Kawasaki H, Mizuseki K, Nishikawa S, Kaneko S, Kuwana Y, Nakanishi S, Ni-shikawa SI, Sasai Y (2000) Induction of midbrain dopaminergic neuronsfrom ES cells by stromal cell-derived inducing activity. Neuron 28:31–40.

Kawauchi T, Chihama K, Nabeshima Y, Hoshino M (2003) The in vivo rolesof STEF/Tiam1, Rac1 and JNK in cortical neuronal migration. EMBO J22:4190 – 4201.

Kim MJ, Woo SJ, Yoon CH, Lee JS, An S, Choi YH, Hwang SG, Yoon G, LeeSJ (2011) Involvement of autophagy in oncogenic K-Ras-induced ma-lignant cell transformation. J Biol Chem 286:12924 –12932.

Kishimoto H, Nakagawa K, Watanabe T, Kitagawa D, Momose H, Seo J,Nishitai G, Shimizu N, Ohata S, Tanemura S, Asaka S, Goto T, Fukushi H,Yoshida H, Suzuki A, Sasaki T, Wada T, Penninger JM, Nishina H, KatadaT (2003) Different properties of SEK1 and MKK7 in dual phosphoryla-tion of stress-induced activated protein kinase SAPK/JNK in embryonicstem cells. J Biol Chem 278:16595–16601.

Koizumi H, Tanaka T, Gleeson JG (2006) Doublecortin-like kinase func-tions with doublecortin to mediate fiber tract decussation and neuronalmigration. Neuron 49:55– 66.

Kuan CY, Yang DD, Samanta Roy DR, Davis RJ, Rakic P, Flavell RA (1999)The Jnk1 and Jnk2 protein kinases are required for regional specific apo-ptosis during early brain development. Neuron 22:667– 676.

Lee S, Sunil N, Shea TB (2011) C-terminal neurofilament phosphorylationfosters neurofilament-neurofilament associations that compete with ax-onal transport. Cytoskeleton (Hoboken) 68:8 –17.

Marek KW, Kurtz LM, Spitzer NC (2010) cJun integrates calcium activityand tlx3 expression to regulate neurotransmitter specification. Nat Neu-rosci 13:944 –950.

McEvilly RJ, de Diaz MO, Schonemann MD, Hooshmand F, Rosenfeld MG(2002) Transcriptional regulation of cortical neuron migration by POUdomain factors. Science 295:1528 –1532.

Nakagawa K, Sugahara M, Yamasaki T, Kajiho H, Takahashi S, Hirayama J,Minami Y, Ohta Y, Watanabe T, Hata Y, Katada T, Nishina H (2010)Filamin associates with stress signalling kinases MKK7 and MKK4 andregulates JNK activation. Biochem J 427:237–245.

Okada S, Nakamura M, Katoh H, Miyao T, Shimazaki T, Ishii K, Yamane J,Yoshimura A, Iwamoto Y, Toyama Y, Okano H (2006) Conditional ab-lation of Stat3 or Socs3 discloses a dual role for reactive astrocytes afterspinal cord injury. Nat Med 12:829 – 834.

Sabapathy K, Jochum W, Hochedlinger K, Chang L, Karin M, Wagner EF(1999) Defective neural tube morphogenesis and altered apoptosis in theabsence of both JNK1 and JNK2. Mech Dev 89:115–124.

Saito T (2006) In vivo electroporation in the embryonic mouse central ner-vous system. Nat Protoc 1:1552–1558.

Sasaki T, Ishiguro K, Hisanaga S (2009) Novel axonal distribution ofneurofilament-H phosphorylated at the glycogen synthase kinase 3beta-phosphorylation site in its E-segment. J Neurosci Res 87:3088 –3097.

Schramek D, Kotsinas A, Meixner A, Wada T, Elling U, Pospisilik JA, NeelyGG, Zwick R-H, Sigl V, Forni G, Serrano M, Gorgoulis VG, Penninger JM(2011) The stress kinase MKK7 couples oncogenic stress to p53 stabilityand tumor suppression. Nat Genet 43:212–219.

Sehara K, Toda T, Iwai L, Wakimoto M, Tanno K, Matsubayashi Y, KawasakiH (2010) Whisker-related axonal patterns and plasticity of layer 2/3neurons in the mouse barrel cortex. J Neurosci 30:3082–3092.

Shimizu S, Konishi A, Nishida Y, Mizuta T, Nishina H, Yamamoto A, Tsu-jimoto Y (2010) Involvement of JNK in the regulation of autophagic celldeath. Oncogene 29:2070 –2082.

Tabata H, Nakajima K (2008) Labeling embryonic mouse central nervoussystem cells by in utero electroporation. Dev Growth Differ 50:507–511.

Takei Y, Teng J, Harada A, Hirokawa N (2000) Defects in axonal elongationand neuronal migration in mice with disrupted tau and map1b genes.J Cell Biol 150:989 –1000.

Toda T, Hayakawa I, Matsubayashi Y, Tanaka K, Ikenaka K, Lu QR, KawasakiH (2008) Termination of lesion-induced plasticity in the mouse barrelcortex in the absence of oligodendrocytes. Mol Cell Neurosci 39:40 – 49.

Ura S, Nishina H, Gotoh Y, Katada T (2007) Activation of the c-JunN-terminal kinase pathway by MST1 is essential and sufficient for theinduction of chromatin condensation during apoptosis. Mol Cell Biol27:5514 –5522.

Wada T, Nakagawa K, Watanabe T, Nishitai G, Seo J, Kishimoto H, Kitagawa D,Sasaki T, Penninger JM, Nishina H, Katada T (2001) Impaired synergisticactivation of stress-activated protein kinase SAPK/JNK in mouse embryonicstem cells lacking SEK1/MKK4: different contribution of SEK2/MKK7 iso-forms to the synergistic activation. J Biol Chem 276:30892–30897.

Wada T, Joza N, Cheng HY, Sasaki T, Kozieradzki I, Bachmaier K, Katada T,Schreiber M, Wagner EF, Nishina H, Penninger JM (2004) MKK7 cou-ples stress signalling to G2/M cell-cycle progression and cellular senes-cence. Nat Cell Biol 6:215–226.

Wang X, Nadarajah B, Robinson AC, McColl BW, Jin JW, Dajas-Bailador F,Boot-Handford RP, Tournier C (2007) Targeted deletion of themitogen-activated protein kinase kinase 4 gene in the nervous systemcauses severe brain developmental defects and premature death. Mol CellBiol 27:7935–7946.

Westerlund N, Zdrojewska J, Padzik A, Komulainen E, Bjorkblom B, Ran-nikko E, Tararuk T, Garcia-Frigola C, Sandholm J, Nguyen L, Kallunki T,Courtney MJ, Coffey ET (2011) Phosphorylation of SCG10/stathmin-2determines multipolar stage exit and neuronal migration rate. Nat Neu-rosci 14:305–313.

Whitmarsh AJ, Cavanagh J, Tournier C, Yasuda J, Davis RJ (1998) A mam-malian scaffold complex that selectively mediates MAP kinase activation.Science 281:1671–1674.

Wolfer DP, Henehan-Beatty A, Stoeckli ET, Sonderegger P, Lipp HP (1994)Distribution of TAG-1/axonin-1 in fibre tracts and migratory streams ofthe developing mouse nervous system. J Comp Neurol 345:1–32.

Xu P, Das M, Reilly J, Davis RJ (2011) JNK regulates FoxO-dependent au-tophagy in neurons. Genes Dev 25:310 –322.

Yasuda J, Whitmarsh AJ, Cavanagh J, Sharma M, Davis RJ (1999) The JIPgroup of mitogen-activated protein kinase scaffold proteins. Mol Cell Biol19:7245–7254.


development/plasticity/repair stress ... · development/plasticity/repair...

Documents